comp.094 final report...southern red meat production – a life cycle assessment page iv the...

final repport

Project code: COMP.094

Prepared by: Greg Peters, Hazel Rowley, Robyn Tucker, Stephen Wiedemann, Michael Short, Matthias Schulz, Andrew Feitz Centre for Water & Waste Technology

Civil & Environmental Engineering

University of NSW; FSA Consulting

Date published: February 2009

ISBN: 9781741915501

Meat & Livestock Australia Limited Locked Bag 991 NORTH SYDNEY NSW 2059

Meat & Livestock Australia acknowledges the matching funds provided by the Australian Government to support the research and development detailed in this publication.

This publication is published by Meat & Livestock Australia Limited ABN 39 081 678 364 (MLA). Care is taken to ensure the accuracy of the information contained in this publication. However MLA cannot accept responsibility for the accuracy or completeness of the information or opinions contained in the publication. You should make your own enquiries before making decisions concerning your interests. Reproduction in whole or in part of this publication is prohibited without prior written consent of MLA.

Southern Red Meat Production – a Life Cycle Assessment


Page ii

Abstract As a major land manager and source of significant greenhouse emissions, the red meat industry could have significant opportunities to enhance the environmental performance of the Australian economy. To achieve optimal environmental outcomes and target management interventions, managers and policy makers need performance information based on best practice data acquisition and analysis. Environmental life cycle assessment (LCA) is an information tool offering an holistic perspective of the environment and the technical system being assessed, and it is for this reason that it is becoming increasingly commonplace in industrial and agricultural management.

Detailed process analysis of farm resource use and productivity was complimented by input-output analysis of service inputs in a hybrid LCA. This report addresses environmental performance indicators including: energy use; global warming potential; solid waste production; eutrophication potential; soil acidification potential and nutrient balances (nitrogen, phosphorus and potassium). The energy and global warming results are comparable with previously published work while the other indicators are not routinely reported for red meat LCAs. While the underlying data for water use are consistent with published results, we demonstrate the influence of different accounting approaches on the results and suggest that approaches that uncritically include rainfall produce counter-intuitive results.

This project enhances the quality of information available to policy makers and others who want to know the answer to questions like: “What is the carbon footprint of red meat?”, “How much energy is used in making red meat?” and “Is much waste produced?” The project also tested an improved suite of agricultural performance indicators for assessing natural resource management issues in LCA.


Page iii

Executive Summary Meat and Livestock Australia (MLA) engaged the Centre for Water and Waste Technology (CWWT) in the School of Civil and Environmental Engineering at UNSW and FSA Consulting to produce an environmental life cycle assessment (LCA) of red meat production. The project investigated three supply chains:

1. a sheep meat supply chain in Western Australia;

2. a premium export beef supply chain in southern NSW; and

3. an organic beef producer in Victoria.

Data were collected for the 2002 and 2004 calendar years for each supply chain. Primary data were obtained on each chain via site visits and reviews of information systems used by the individual property managers. Environmental indicators were selected at a workshop of project stakeholders including MLA officers and red meat producers. This report presents the results of the LCA work undertaken using these data, and discusses associated methodological issues.

This study is the first detailed LCA of red meat production in Australia. It combines detailed process-based LCA with high level input-output analysis to present a more accurate and complete picture of the environmental profile of supply chains than is feasible using either process LCA or input-output analysis in isolation.

The global warming potential of the three supply chains ranged from 6.8 to 11 kg CO2-e per kg HSCW. The highest value was for the organic beef supply chain and the lowest was for the sheep meat supply chain. The presence of a feedlot in the premium export beef supply chain reduced that supply chain’s greenhouse emissions – the lower enteric methane emissions resulting from more efficient feed conversion outweighed the additional carbon dioxide emissions associated with feed production.

Energy use varied between 24 and 30 MJ per kg HSCW. The meat processing facility generally constituted most of the energy demand in all three supply chains. Diesel consumption for stock transport did not contribute significantly to the total figures.

Estimates of water use that include rain used to provide drinking water, grow fodder and other feedstuffs for red meat production range from 15,000 to 105,000 L/kg in the literature1. When we include rain our estimates range from 7,387 to 57,634 L/kg HSCW post processor depending on the supply system and production year. When rain is excluded but significant irrigation occurs, the amount of water use estimated in the literature drops to a value in the thousands of litres. Most Australian red meat production does not involve significant irrigation. Without significant irrigation or rainfall included in the calculus, average water use falls to the hundreds of litres per kilogram. Our estimates range from 18 to 540 L/kg (the higher figures reflecting the production of irrigated feedsstuffs in our NSW example) and are consistent with the literature data.

1 Note that it is not always clear from the reports as to stage of the supply chain at which the masses are computed, nor whether it is expressed as boneless.


Page iv

The eutrophication potential exhibited a large range from 0.13 to 1.2 kg O2 depletion per kg HSCW. Low estimated soil erosion at the grazing property helped the sheep meat supply chain outperform the other two chains for this indicator.

Reliable data on solid waste generation were particularly hard to obtain. We produced some estimates using two bases – including and excluding the organic wastes produced at the meat processing works. The former estimates range from 0.042 to 0.065 kg/kg HSCW. If the organic wastes are excluded, the range contracts to 0.021 to 0.039 kg/kg HSCW.

This work advanced the methodological development of life cycle impact assessment (LCIA) by examining the feasibility of novel indicators for natural resource management issues relevant to agricultural LCA. Where existing methodologies were followed the results are consistent with other work in agricultural LCA. Where new indicators were developed, this project presents results that can be benchmarked against other production systems as the application of these indicators progresses. It also offers insights into the variability of the three case study supply chains across different regions of Australia.

The nutrient management indicators suggested that the nitrogen (N) account for the grazing properties varied from a 0.028 kg N per kg HSCW loss to a 0.17 kg per kg HSCW accumulation of N on farm. The main contributors to these changes are growth of N-fixing pastures (or lack thereof) and the application of fertilisers.

The sheep and premium export supply chains also accumulated between 0.0085 and 0.019 kg phosphorus (P) per kg HSCW. Losses of 0.0039-0.0051 kg P per kg HSCW in the organic beef supply chain reflect a strategic decision by the property manager. This manager also uses mineral additives to significantly increase potassium (K), resulting in an accumulation of 0.095 kg K per kg HSCW. This is compared with absolute values at least a factor of four lower for the other supply chains.

This management activity is also reflected in the soil acidification indicator, which in 2002 showed a farm surplus of 630 kg CaCO3-equivalent per hectare and year, while the other supply chains all showed a deficit of less than 23 kg CaCO3-e/ha.y. In this report we argue that soil acidification, and the soil erosion potential indicator, are best described on an area basis rather than by the kg HSCW produced, although both results are shown in the report. Soil erosion potential also varied across the three supply chains, with the NSW chain exhibiting the highest erosion potential due to the characteristics of the soils and topography.


Page v

Contents

1 Introduction................................................................. 1 1.1 Project Description ........................................................................1

1.2 Background ....................................................................................1

1.3 Meat and related LCAs ..................................................................2

1.4 Australian Dairy LCA .....................................................................4

2 Life Cycle Assessment Methodology....................... 5 2.1 Overview .........................................................................................5

2.2 Expanding the LCI using Input-Output Analysis.........................6

2.3 Life Cycle Impact Assessment – Suitability of Impact Models 7

2.4 Efficient Water Use ........................................................................7

2.5 Energy / Greenhouse .....................................................................8

2.6 Solid Waste.....................................................................................9

2.7 Nutrient Management ....................................................................9

2.8 Soil Acidification Potential ..........................................................16

2.9 Soil Erosion ..................................................................................22

2.10 Water Quality................................................................................22

3 Goal and Scope Definition....................................... 25 3.1 Goal ...............................................................................................25

3.2 Scope of the Study.......................................................................26

3.3 LCA Model ....................................................................................29

4 Life Cycle Inventory ................................................. 30 4.1 Introduction ..................................................................................30

4.2 Grazing Properties .......................................................................30


Page vi

4.3 Lot Feeding Property ...................................................................32

4.4 Meat Processing...........................................................................33

4.5 Inputs to the Agricultural System...............................................34

4.6 Outputs from the Agricultural Production System ...................42

4.7 Expanded Supply Chain (Input-Output Analysis) .....................42

4.8 Relating LCI Data to the Functional Unit....................................43

5 Life Cycle Impact Assessment: results and discussion................................................................. 50

5.1 Conventional LCIA Indicators.....................................................50

5.2 Results for Natural Resource Management Indicators.............70

5.3 Sensitivity Analyses ....................................................................76

5.4 Comparison with published literature........................................77

6 Conclusions.............................................................. 84

7 Recommendations ................................................... 85

8 Bibliography.............................................................. 86

9 Appendices ............................................................... 94


Page vii

Table of Figures Figure 1: General framework for LCA and its application (ISO 14040, 1999) ........5

Figure 2: General LCA system model of red meat sector ....................................28

Figure 3: Process flow chart for a typical meat plant showing inputs and outputs (MLA 2002, p 116)..................................................................34

Figure 4: Theoretical mass balance for a beef processing plant (MLA 2002, p 89)....................................................................................................49

Figure 5: GWP contributions in the entire supply chain (by stage).......................51

Figure 6: GWP contributions in the entire supply chain (by activity) ....................52

Figure 7: GWP contributions of activities at each of the grazing properties .........53

Figure 8: GWP contributions of activities at the feedlot (NSW) ............................54

Figure 9: GWP contributions of activities in meat processing ..............................54

Figure 10: Primary energy use in the entire supply chain (by stage) ...................55

Figure 11: Primary energy use in the entire supply chain (by activity) .................56

Figure 12: Primary energy use of activities at each of the grazing properties ......57

Figure 13: Primary energy use of activities at the feedlot (NSW).........................58

Figure 14: Primary energy use of activities in meat processing ...........................58

Figure 15: Eutrophication potential contribution in the entire supply chain (by stage) ............................................................................................60

Figure 16: Eutrophication potential contribution in the entire supply chain (by activity) ..........................................................................................60

Figure 17: Eutrophication potential of activities at each of the grazing properties ............................................................................................62

Figure 18: Eutrophication potential of activities at the feedlot (NSW) ..................62

Figure 19: Eutrophication potential of activities in meat processing.....................63

Figure 20: Solid waste generation in the entire supply chain (by stage) ..............64

Figure 21: Solid waste generation in the entire supply chain (by activity) ............65

Figure 22: Solid waste generation by activities at each of the grazing properties ............................................................................................66

Figure 23: Solid waste generation by activities at the feedlot (NSW)...................67


Page viii

Figure 24: Solid waste generation by activities in meat processing .....................68

Figure 25: Solid waste generation (excluding manure+paunch) in the entire supply chain (by stage) .......................................................................69

Figure 26: Solid waste generation (excluding manure+paunch) in the entire supply chain (by activity) .....................................................................69

Figure 27: Solid waste generation (excluding manure+paunch) by activities in meat processing ..............................................................................70

Figure 28: N balance for the grazing properties ...................................................71

Figure 29: P balance for the grazing properties ...................................................72

Figure 30: K balance for the grazing properties ...................................................73

Figure 31: Soil acidification at the grazing properties (per kg HSCW)..................74

Figure 32: Soil acidification at the grazing properties (per ha.yr) .........................74

Figure 33: Soil erosion at the grazing properties (per ha.yr) ................................75

Figure 34: GWP for beef and lamb production (unallocated farm gate kg CO2-e / kg HSCW)..............................................................................80

Figure 35: Primary energy (unallocated farm gate MJ / kg HSCW) .....................81


Page ix

Table of Tables Table 1: Summary of environmental issues of concern for MLA ...................... 7

Table 2: N fixation from legume pastures as cited in the literature................. 11

Table 3: Summary of nutrient input values and assumptions used for properties in the red meat supply chains ........................................... 12

Table 4: Nitrate leaching rates under clover based pastures in southern Australia ............................................................................................ 14

Table 5: Volatilisation and denitrification losses from agricultural systems in Australia ............................................................................................ 15

Table 6: Summary of nutrient output values and assumptions used for properties in the red meat supply chains ........................................... 16

Table 7: Lime required to neutralise the acidifying effects of some nitrogenous fertilisers at different rates of NO3 leaching.................... 17

Table 8: Potential acidification from sheep grazing behaviour ....................... 18

Table 9: Alkalinity in exported agricultural produce and lime requirement to neutralise acidifying effect of product removal................................... 20

Table 10: Summary of acidification potential data used for properties in the supply chains..................................................................................... 21

Table 11: Nutrient losses from pasture systems is Australia.......................... 23

Table 12: Greenhouse gas emission estimates for the grazing properties..... 31

Table 13: Greenhouse gas emission estimates for the lot feeding property... 32

Table 14: Resource use and waste generation data for a typical meat processing plant (UNEP Working Group for Cleaner Production, cited in MLA 2002, p 4) ..................................................................... 35

Table 15: Fuel combustion emission factors for coal (stationary energy)....... 37

Table 16: LCI data for grains production (after Narayanaswamy et al. 2004) 39

Table 17: LCI data for hay production (after Cederberg 1998)....................... 40

Table 18: ‘Depths’ of production orders included in the process analysis ...... 43

Table 19 – Livestock classes and growth rates for the cattle supply chain properties .......................................................................................... 44

Table 20 – Livestock classes and growth rates for the sheep supply chain property ............................................................................................. 45


Page x

Table 21: Allocation data for process units .................................................... 48

Table 22 – Water use (should not be cited without reference to the full report)................................................................................................ 59

Table 22: Comparison of grass and grain finished beef ................................. 76

Table 23: Parameter sensitivity analysis of the NSW supply chain in 2004 ... 77

Table 25: Dressing percentage and saleable meat percentage from literature ............................................................................................ 79


Page xi

List of Abbreviations

a annum

ABS Australian Bureau of Statistics

AGO Australian Greenhouse Office

CO2-e Carbon dioxide equivalents

DRDC Dairy Research and Development Corporation

GWP Global Warming Potential

HSCW Hot Standard Carcase Weight

IOA Input-Output Analysis

IPCC Intergovernmental Panel on Climate Change

ISO International Organization for Standardization

LCA Life Cycle Assessment

LCI Life Cycle Inventory

LCIA Life Cycle Impact Assessment

NGGIC National Greenhouse Gas Inventory Committee

NLWRA National Land & Water Resources Audit

UNEP United Nations Environment Programme


Page 1

1 Introduction

1.1 Project Description

Meat and Livestock Australia (MLA) engaged the Centre for Water and Waste Technology and FSA Consulting to perform an environmental life cycle assessment (LCA) of red meat production in Australia. The aim was to obtain relatively detailed (process-based) life cycle inventory (LCI) data for several supply chains.

The first phase of the project was a literature review of LCA data and approaches relevant to red meat production. This review is briefly summarised in this report.

The second phase involved LCI compilation. The project team has successfully liaised with three supply chains representing different parts of southern Australia: a sheep meat supply chain in Western Australia; a sheep and cattle based supply chain in southern NSW and an organic beef producer in Victoria. Data were collected for the 2002 and 2004 calendar years for each of these three supply chains. We would like to extend this work by considering at least one northern supply chain.

The third phase was a dialogue with project stakeholders regarding priorities in life cycle indicator enhancement. This is summarised in Section 2.2 of this report.

The fourth phase was analysis of the LCA results using a combination of the GaBi software package and CWWT’s proprietary input-output analytical model. This report provides detailed analysis of the model results and considers some of the methodological issues encountered at various points in the study. It will be the basis for scientific publication of the results.

1.2 Background

The red meat industry is one of Australia’s largest agricultural industries, with a gross value of production in excess of $9.5 billion (2006/2007). Australia is the second largest beef exporter and second largest sheep meat exporter in the world (MLA 2007; MLA 2007). The Australian red meat industry, like many other Australian primary industries, is coming under increasing pressure from both the community and government to document and justify its impact on the environment. Environmental management will also be important at an enterprise level in the future as it is likely to play a major role in determining competitive advantage, especially in export markets.

MLA has commissioned many research projects over the last decade to improve the environmental performance of the red meat industry. The red meat processing sector, in particular, has been the subject of intensive environmental research aimed at improving factory wastewater treatment, benchmarking environmental performance and quantifying greenhouse gas emissions. Many of the research outcomes and recommendations have been adopted by the industry through programmes like the Environmental Management Systems Manual and Eco-Efficiency Manual (MLA 2002). Research in the grazing sector has focussed primarily on improving profitability, productivity and sustainability. The


Page 2

results from the highly successful Sustainable Grazing System (SGS) program have been widely disseminated throughout the industry (MLA 2003) and the scientific literature (e.g. Aust. J. Exp. Ag. (2003) 43). Testing the SGS in a national experiment showed it to be more profitable and sustainable than other grazing systems, with improved pasture composition and persistence, increased ground cover, lower acidification, salinity, erosion, better water quality and increased biodiversity (MLA 2003). SGS and its subsequent education programmes (e.g. PROGRAZE) form the basis of MLA’s current EDGEnetwork workshops. These workshops aim to improve producers understanding of the management of grazing, fertiliser applications, soils, water resources, pastures, weeds and biodiversity. MLA has also funded a significant amount of environmental research for the lot feeding sector (e.g. FLOT.132 – 2020 Vision of the beef industry; FLOT.328; Environmental Sustainability Assessment of the Australian Feedlot Industry).

While both the livestock production and processing sectors are achieving environmental successes, there is an absence of data on the environmental impacts of the red meat industry as a whole. The red meat production and processing industry are only two processes in the supply chain for the delivery of red meat to domestic and export markets. Without quantification of the impacts of dependent industries – for example, feed supplement industries such as grain, hay, molasses and oilseed/protein meal; transportation; fertiliser; pesticides and herbicides; energy and packaging – the industry is not in a position to optimise the environmental impacts of the entire system. Sectors of the supply chain beyond livestock production and meat processing, like transportation, feed and energy could play a significant role in the overall environmental impact. A whole of life cycle view offers the potential to identify areas where gains will be possible and hence the opportunity to help bring about an overall improvement in industry environmental (and business) performance. For example, there may be opportunities for producers to collectively improve resources management through engagement with suppliers, e.g. feed, fertiliser and transportation. Also, the industry can avoid creating new environmental problems through a greater understanding of their whole system. This is particularly valuable since “solving” an environmental problem in one sector (e.g. grazing property) to the detriment of another sector (e.g. feed supplementation) may not be sustainable in the long term. There may be opportunities for processors to improve their environmental performance through product stewardship/supply chain management, i.e. red meat supply, energy, packaging, transport and distribution.

1.3 Meat and related LCAs

Consumers are beginning to make product selection choices on the basis of environmental considerations and the environment is an area of potential non-tariff trade barriers to the international market. MLA’s marketing of Australian lamb and beef in the USA, Japan and the Middle East is based on a “clean, green and safe” image. With increasing competition for export markets, it is likely that the industry will be called on to quantitatively justify its green image at some time in the future. LCA is a useful environmental tool for this purpose as it can quantify the environmental impacts of an entire industry. LCA has already been used by governments in their decisions on the development of industry legislation and will continue to be used in the future. Examples are several European Directives on packaging material, chemicals, waste management sector and take back schemes for automotives.


Page 3

The importance of LCA for meat products is recognised in Europe and Japan with LCAs having been conducted on pork (Basset-Mens and van der Werf 2005; Eriksson 2005), lamb (Schlich and Fleissner 2005) and beef (Cederberg and Mattsson 2000; Haas et al. 2001; Ogino et al. 2004; Chassot et al. 2005). Related LCAs on leather (Canals et al., 2002) and particularly milk and dairy products (Blonk and van Zeijts 1997; Cederberg and Mattsson 2000; Haas et al. 2001; Berlin 2002; Eide 2002; Svenskmjölk 2002; de Boer 2003; Hospido et al. 2003; Lundie et al. 2003; Casey and Holden 2005) have been the subject of intense investigation. Beef in northern Europe is primarily sourced from dairy cattle and many studies grapple with the complexity of allocating environmental impacts to milk, meat and other co-products such as leather. In all cases, feed choice and feed production systems are the major contributors to the environmental impact.

Grains are an important feed component in the lot feeding sector of the meat industry. While most grains are produced from dryland crops in Australia rather than using significant irrigation resources, there is a substantial potential ecotoxicity burden associated with the use of pesticides and herbicides during production and storage (Narayanaswamy et al. 2005). In the Australian Dairy LCA, feed supplementation with grains represented a primary source of ecotoxicity in milk and thus in grinding beef for export to the USA. A recent Swiss study by Chassot et al. (2005) compared pasture to lot feeding beef and concluded that the differences between the two systems were relatively minor except for the considerably greater ecotoxicity impacts for the feedlot system resulting from greater fertiliser use.

Oilseeds and protein meals are important feed supplements in the lot feeding sector, with whole cottonseed, cottonseed meal and other protein meals frequently included in feedlot cattle diets. These materials are also commonly used in drought feeding situations on farms. An LCA of rapeseed oil production for use as a chainsaw lubricant has been conducted in the UK (Wightman et al. 1999) and, more recently, Australian LCI data for canola oil production were compiled by Narayanaswamy et al. (2004).

Wood et al. (2007) examined organic and conventional farming practices in Australia using a hybrid input-output LCA methodology. Their examination of meat, grain, fruit and vegetable production showed that while the direct on-site use of energy and materials of the organic farms exceeded those of the conventional farms, organic farming performed better when the entire supply chain was considered, except in the case of sheep and wheat production.

As part of an MLA project FLOT.328 (“Environmental Sustainability Assessment of the Australian Feedlot Industry”) conducted in parallel to this project, the research team undertook a detailed review of literature on water use in feedlot processes (FSA Consulting 2005). This study identified a number of information sources that can be used to enhance the LCI analysis of water use on cattle grazing properties, including research by Winchester & Morris (1956), Hicks et al. (1988), Sanders et al. (1994) and Parker et al. (2000).

A similar review was undertaken on energy use and greenhouse gas emissions from Australian feedlot operations (FSA Consulting 2005) also as part of the FLOT.328 project. This study identified information sources that can be used to enhance the LCI analysis of energy use and greenhouse gas emissions of cattle grazing properties, including research by Lipper et al. (1976), Sweeten & McDonald (1979), Schake et al. (1981), Sweeten et al. (1986), Casada & Safley (1990), Sweeten (1990), Safley et al. (1992), Johnson & Johnson (1995), Steed & Hashimoto (1995), IPPC (1997), Harper et al.


Page 4

(1999), Hegarty (1999), Hao et al. (2001), Hegarty (2001), Woodbury et al. (2001), NGGIC (2002), Page (2003), Tedeschi et al. (2003), AGO (2004), NGGIC (2004), ABS (2005), McGrabb (2005) and QDPI&F (2005).

1.4 Australian Dairy LCA

In Australia, both Dairy Australia and the Grains Research and Development Corporation have made significant investment in life cycle analysis and are using the results to target environmental improvements in their respective industries (Nicol 2005). Small individual Australian dairy farm case studies have found that the environmental impacts associated with the provision of feed are substantial (Wegener 1999; Chen et al. 2005). A preliminary Australian meat LCA was conducted for the Cannon Hill meat processing facility (Gibson 2002; Renouf 2002). The work primarily quantified water use, energy use and greenhouse gases for the processing site. Sugar production and electricity generation from sugarcane bagasse was also studied using LCA (Renouf 2002; Renouf 2002).

The Australian Dairy LCA has direct relevance to MLA as it forms the primary LCI for the US export grinding beef market. The dairy farms located in southern Australia fall into three major groups (DRDC 2001):

Murray Dairy: >90% irrigation, 1.4 t feed/cow/yr, 4,700 L milk/cow/yr.

Gippsland Dairy: ~50% irrigation, 0.8 t feed/cow/yr, 4,600 L milk/cow/yr.

Western Victoria:


Page 5

2 Life Cycle Assessment Methodology

2.1 Overview

Life Cycle Assessment (LCA) is a form of cradle-to-grave systems analysis developed for use in manufacturing and processing industries to assess the environmental impacts of products, processes and activities by quantifying their environmental effects throughout the entire life cycle. LCA can be used to compare alternative products, processes or services; compare alternative life cycles for a product or service; and identify those parts of the life cycle where the greatest improvements can be made. An international standard has now been developed to specify the general framework, principles and requirements for conducting and reporting LCA studies (Blamey et al. 1998). LCA differs from other environmental tools (e.g. risk assessment, environmental performance evaluation, environmental auditing, and environmental impact assessment) in a number of significant ways. In LCA, the environmental impact of a product or the function a product is designed to perform is assessed, the data obtained are independent of any ideology and it is much more complex than other environmental tools (UNEP 1996). As a system analysis, it surpasses the purely local effects of a decision and indicates the overall effects.

Direct application:

- Product developmentand improvement

- Strategic planning

- Public policy making

- Marketing

Goal and scopedefinition

InterpretationInventory analysis

ImpactAssessment

Life Cycle Assessment Methodology

Figure 1: General framework for LCA and its application (ISO 14040, 1999)

There are four phases of LCA:


Page 6

Goal and Scope Definition defines the goal, functional unit and associated system to be studied.

Inventory Analysis analyses all process inputs and outputs. It involves modelling unit processes in the system, considered as inputs from the environment (resources, energy) and outputs (product, emissions, waste) to the environment. Allocation of inputs and outputs needs to be clarified where processes have several functions (for example, one production plant produces several products). In this case, different process inputs and outputs are attributed to different goods and services produced. An extra simplification used by LCA is that processes are generally described without regard to their specific location and time of operation.

Impact Assessment makes results from the inventory analysis more manageable and understandable in relation to natural environment, human health and resource availability.

Interpretation involves evaluating inventory analysis and impact assessment outcomes against the study’s goal.

An LCA is essentially a quantitative study. However, not all environmental impacts can be quantified due to a lack of data or inadequate impact assessment models. A guide to decisions can then be made through qualitative use of LCA and other tools for supply chain analysis. Quantitative analysis requires standardised databases of main processes (energy, transport) and software for managing the study’s complexity.

2.2 Expanding the LCI using Input-Output Analysis

Input-output analysis (IOA) is a mathematical modelling technique based on a model of the national economy that can be used to ‘fill in the gaps’ of an LCA when detailed process-based engineering LCI data are unavailable.

The production systems in this LCA were studied in detail to include as much physical data as possible (as presented in previous sections). However, it is never possible to include the entire system. For example, how much water is consumed during the manufacturing of vehicle repair materials for the feedlot? Collecting such detailed data would be impractical and expensive. Therefore, LCA practitioners generally identify a system boundary based on experience in LCA and dialogue with the owners or managers of the system under study. Depending on the topic and purpose of the LCA, this may not be a problem, but in some cases it can lead to underestimation of the material or energy budget of a production system.

IOA has been proposed as an alternative to conventional LCA, because it overcomes these limitations. IOA involves constructing a mathematical model of the national economy and the environmental impacts of industries. The model can be used to estimate the environmental impacts of any producer based on that producer’s expenditure patterns. However, IOA would not be as accurate as LCA in describing on-farm impacts.

Recent research in this area shows that the most accurate results can be achieved by combining the two techniques – using the precision of LCA to get a detailed picture of the


Page 7

main industry being examined, and using IOA to ‘fill in the gaps’ regarding some of the supporting industries. The research team at CWWT developed a sophisticated hybrid model to improve the accuracy of LCA by incorporating IOA results into it.

2.3 Life Cycle Impact Assessment – Suitability of Impact Models

At the start of this project, MLA identified a number of natural resource management issues of concern to the red meat industry, ranging from energy efficiency to feral animals (Table 1). Some of the natural resource management (NRM) issues can be modelled using the default list of LCA impact categories (de Haes et al. 1999; Guinee 2002). These include water quality, water use efficiency, eutrophication, energy use efficiency and greenhouse gas emissions and solid waste. However, conventional LCA impact models do not adequately cover the remaining NRM issues.

A workshop was held on the 18th of August 2006 to engage project stakeholders in a process of information sharing and prioritisation of NRM issues for this project. Considering the available models and the significance of the issues, the workshop prioritised the items in Table 1 marked with an asterisk.

Strategies for addressing these nonconforming NRM issues within a LCA framework are outlined below.

Table 1: Summary of environmental issues of concern for MLA (* denotes issues able to be modelled using conventional LCA input categories)

NRM Issues of concern

Water quality*

Water use efficiency*

Salinity

Soil erosion*

Nutrient management*

Soil acidification*

Weeds

Feral animals

Biodiversity

Vegetation management

Energy efficiency & greenhouse gas emissions*

Solid Waste*

2.4 Efficient Water Use

On livestock grazing properties and feedlots, water uses include irrigation, stock drinking, feed processing, cattle washing and trough cleaning. Clearing land for grazing may increase runoff from properties, while installing small agricultural dams may reduce it. In


Page 8

this work, water use is defined consistent with the definition used by the Australian Bureau of Statistics (ABS 2004):

“… water extracted directly from the environment for use, [which] includes water from rivers, lakes, farm dams, groundwater and other water bodies. Some of this water is then distributed via a water provider to other water users. The volume of water used from rainfall is not in scope of the water account, unless it was stored and/or delivered before use. For example, rainfall directly onto a crop is not in scope for the water account. However, if rainfall is collected in a farm dam and then applied onto the crop, it is in scope and is included in the self-extracted water use figures.’

This is consistent with the work of various other strategic environmental assessments in Australia (e.g. Foran et al. 2005) and overseas (e.g. Beckett and Oltjen 1993). Water is considered “used” if it is either transferred from its natural watercourse or extracted from underground aquifers. By definition, dryland cropping does not “use” water. Similarly, rain that collects in a small agricultural dam within the property and consumed in situ is not “used” unless it is pumped to a location outside the catchment of that dam.

On-farm water use was estimated in this project using data supplied by individual producers. A farm hydrological water balance was constructed to account for all water inputs and outputs. In addition, water use by suppliers of goods and services to the grazing property, feedlot and meat processing works was estimated during life cycle modelling. Water use was reported on in the previous report which is included as Appendix E.

2.5 Energy / Greenhouse

Energy consumption was estimated in this project from data supplied by individual producers. Data from the Australian Greenhouse Office (AGO 2004) were used to convert raw energy data (e.g. litres of diesel) into primary energy consumption (e.g. megajoules of primary energy). Primary energy is also referred to as “full cycle” energy and means, for example, that electricity consumption is not only related back to the coal burnt to generate it, but to the energy involved in obtaining the coal. Primary energy consumption is then compared with the production of beef in kg HSCW.

Greenhouse gas emissions were estimated on the basis of the energy consumption data obtained in the LCI phase of the project. Emissions due to livestock transport, commodity delivery, water supply, administration and effluent irrigation sectors are included in the life cycle impact assessment (LCIA) and the total data are normalised against kg HSCW production. Emissions from agricultural livestock production are, in this study and generally, calculated by multiplying estimates of activity levels (such as cattle numbers, diet composition and manure production) by emission factors drawn from the National Greenhouse Gas Inventory Committee (NGGIC 2004).

Global warming potential (GWP) is usually evaluated on a 20, 100 or 500 year timescale. For this study, the most commonly used timescale was selected - 100 years. The relative contributions of each greenhouse gas to GWP were estimated by using equivalence factors set in the most recent publication by the Australian Government (DCC 2008) rather than the latest from the Intergovernmental Panel on Climate Change (IPCC).


Page 9

2.6 Solid Waste

Solid wastes generated on livestock grazing properties include tyres; chemical containers and drums; end-of-life vehicles and equipment; and organic waste (e.g. carcases, spoiled feed). Solid wastes are often disposed of in on-farm tips. Waste production by suppliers of goods and services was also included in the overall analysis. Manure was not considered a waste at the feedlot, where it is reused as a matter of course and does not leave the LCA system boundary. The data for the meat processing works are presented in two ways: considering the paunch and yard manure as wastes, and excluding it from the definition of waste.

In this LCA, wastes were assessed primarily from a resource use perspective. Waste production was estimated from the data supplied by individual producers and feedlot operators and, for meat processing works, from MLA (2002) data. The resources used to produce the waste materials, and the environmental impacts of that production, were incorporated in the life-cycle modelling. Waste is presented as the mass of solid waste produced per kg HSCW produced.

2.7 Nutrient Management

The nutrient balance was calculated using mass balance principles to estimate the nitrogen (N), phosphorus (P) and potassium (K) in major system inputs (incoming livestock, fertiliser, feedstuffs and other nutrient inputs) and outputs (e.g. outgoing livestock, wool and harvested product (e.g. hay, grain)) on an annual basis. The mass of each input or output category was, where possible, calculated from the farm records kept for each property. In the case of inputs and outputs less readily quantified in situ (N fixation by legume pastures, N leaching through the soil profile) estimates from the literature were used to calculate the values used in the balance. The data for nutrient inputs and outputs were collected during the initial grazing property surveys and from the literature. The assumptions used to calculate the N, P and K for the nutrient balance are described below, and a summary of the data is provided in Table 3 and Table 6 at the end of each section.

2.7.1 Nutrient inputs

Livestock

The mass of livestock imported annually onto each property was calculated from farm records kept for the 2002 and 2004 calendar years. The chemical composition for beef entering and exiting the property was estimated using the Beef-bal program (DPI&F 2003). The use of a single number for the composition of livestock introduces the possibility of error for two reasons: 1) differences that occur in body composition for store animals (animals that have a low amount of body fat) compared to finished animals which are ready for sale and typically have a higher percentage of body fat, and 2) variations in body composition between animals of different sexes and ages. Considering the broad property scale nature of this research it was decided that a single figure for all classes of animal would be sufficient because the overall impact is likely to be relatively low compared to sources of error for other components in the nutrient balance.


Page 10

For sheep production, data for live animal composition were sourced from Cornell University (CNMSP 2007). However these data showed less than 5% variation from the estimate used for beef cattle so it was assumed that sheep and cattle were equivalent in terms of nutrient removal from the system per kilogram of liveweight. Wool is a secondary (in terms of mass) output from the enterprise and was accounted for in the nutrient balance by assuming that clean wool weight (assumed at 70% of greasy wool weight) is 100% protein and therefore contains approximately 16% N, with negligible amounts of P and K.

Fertiliser

The mass of fertiliser applied on each supply chain property was collected from farm records for the calendar years 2002 and 2004. All values were calculated as property-scale averages. The composition (chemical analysis) of specific types of fertiliser can vary between manufacturers, but this is usually by less than 1% of N, P or K content. The chemical analysis for fertilisers applied were taken from Incitec (Incitec Pivot 2005; 2005; 2005). The chemical analysis for the organic fertilisers and soil amendments used on the Victorian property were collected from the relevant manufacturer (Nutri-tech 2006). The amount of nutrient used was calculated on an ‘as applied’ basis. Fertiliser usage for the cropping system was included in the nutrient budget so that impacts from grain production exported off-farm (separate to the red meat system) could be accounted for by an allocation factor applied in the final analysis.

Feed and feed supplements

Significant amounts of nutrients can be brought on farm in feed for livestock. The data supplied by producers in the survey for feed/feed supplement purchases were combined with standard figures for dry matter percentage and nutrient composition to estimate nutrient inputs. The nutrient composition for a wide range of commodities was collected for the related research project FLOT.328 and these data provided input for the current project.

Legumes - nitrogen fixation

N fixation is the single most significant nutrient addition not directly derived from grazing property inputs. It was not possible to directly assess the mass of N fixed by legumes on the supply chain properties so data were sourced from a wide breadth of literature for N fixation by legume pastures in Australia. The rate of N fixation reported in the literature varies from 11 - 12 kg N / ha / yr for white clover (8% of pasture biomass) in south western Victoria (Riffkin et al. 1999; McKenzie et al. 2003) to 162 kg N / ha / yr for subterranean clover in a mixed sward in Western Australia (Asseng et al. 1997). However, for mixed clover based pasture swards in southern Australia, the likely rate of N fixation is in the range 40 - 150 kg N / ha / yr (Asseng et al. 1997; Unkovich 1997; Dear et al. 1999; Riffkin et al. 1999). These references are summarised in Table 2.

N fixation is influenced by many factors, the primary drivers being legume species, pasture growth rate and the percentage of legume in the pasture sward. These factors were researched for each of the supply chains to allow a more accurate estimation of the likely rate of N fixation from the legume component. These were then matched as closely as possible to the available literature to determine a rate of N fixation at the property scale. Table 2 gives the N fixation rates that were used to estimate appropriate fixation rates for properties in each of the red meat supply chains.


Page 11

Table 2: N fixation from legume pastures as cited in the literature

Species Grazing conditions

Percentage of pasture

sward Region and rainfall

N fixed

kg / ha / yr Reference

White clover / ryegrass

Not known 12 – 23 %

South western VIC – 790mm 12 – 42

(McKenzie et al. 2003)

Sub clover Intensive grazing Not known Southern Australia 92

(0 – 188) (Unkovich

1997)

Sub clover / Phalaris

Intensive grazing Not known

South western NSW – 575mm 48 – 59

(Dear et al. 1999)

White clover Not known 8 % Not known 11 – 18 (Riffkin et al. 1999)

Sub clover / ryegrasss


South western WA – 673mm in year of

experiment 188 (Sanford et al. 1995)

Sub clover / ryegrasss


South western WA – 673mm in year of

experiment 103 (Sanford et al. 1995)

An estimate of 46 or 92 kg N / ha / yr depending on rainfall in the survey year was assigned for the properties in Western Australia, New South Wales and Victoria that have clover based pastures after Unkovich et al. (1997). Considering all the references in Table 2, it is accepted that the actual N fixation could vary on a yearly basis from 10 - 190 kg N / ha / yr or more depending on the range of driving factors that are difficult to account for. Hence an intermediate figure was selected to try to represent a mean value.

Inputs of N from the atmosphere with rainfall and lightning were included in the mass balance using a single figure for the whole nation, though this may vary from one area to another. It was assumed that this difference is not likely to have a large effect on the mass balance considering the overall addition of N from this source is only around 5 kg N / ha / yr (Sharma and Campbell 2003). Other additions from the atmosphere can include calcium, magnesium and sodium from sea salts which are added at a rate dependent on the prevailing winds and distance from the ocean. There were insufficient data available for the properties involved to include a value for the deposition of these nutrients from the air and hence they have been ignored in this analysis.

Table 3 below shows the values used to calculate the nutrient budget for farms in the three supply chains. Data have been presented as a range covering the three properties and two years of data collection.


Page 12

Table 3: Summary of nutrient input values and assumptions used for properties in the red meat supply chains

Nutrient inputs Description Value Reference

Livestock Sheep / Cattle N = 2.4% of liveweight

P = 0.7% of liveweight

K = 0.2% of liveweight

QDPI&F (2005)

Fertiliser urea,

mono-ammonium phosphate,

di-ammonium phosphate,

sulfate of ammonia

Pivot 15

Organic humates

Seachange kelp mix

single superphosphate

muriate of potash

K sulphate

K humate

Dependant on fertiliser type – taken from manufacturers reported nutrient levels

Incitec Pivot (2005a,b,c)

Feed and feed supplements

Pasture hay

Legume hay

Lupins

Canola meal

Canola oil

Minerals

N=1.3 %, P=0.4%, K=0.2%

N=2.1 %, P=0.4%, K=1.0%

N=4.6 %, P=0.3%, K=0.8%

N=5.8 %, P=1.0%, K=1.2%

N=0.0 %, P=0.3%, K=0.4%

Dependant on type

QDPI&F (2005)

and from manufacturers (mineral supplements).

Legumes - N fixation

Mixed clover / grass pastures

46 kg/ha or 92kg/ha Estimate dependant on density of legume within the pasture stand and the annual rainfall in the survey year

Unkovich et al. (1997) and references in Table 2

2.7.2 Nutrient outputs

Nutrient outputs have two main forms: export of produce and losses to the environment. The nutrient budget first assesses the nutrients added to the system less the nutrients exported in produce. The nutrient balance also defines the pool of nutrients that may be


Page 13

lost to the environment. This provides data to other indicators in the assessment including acidification and water quality.

Quantifying nutrient losses from the system is an important outcome from the nutrient budget. Nutrient loss is a major source of environmental concern, particularly in respect to N and P movement to surface and groundwater sources. Categories for soil/water losses (including nutrient transport with overland flow and erosion), leaching and volatilisation losses were also estimated. Other specific nutrient loss pathways (e.g. fire) were considered minor pathways in these southern meat supply chains and excluded from this assessment. All of the loss pathways identified in the assessment were estimated from an appraisal of the literature and the system under analysis. The data used in the modelling are presented in Table 6. It was beyond the scope of the project to measure nutrient losses for the specific properties on site.

Export of produce

Exports include animals for slaughter, wool and other produce. Other produce (i.e. grain) was excluded from the mass balance except where directly linked to red meat production. Wool production could not easily be excluded from the mass balance, introducing a degree of complexity. Where wool was produced, the impact categories were adjusted to apportion impacts to wool and meat on the basis of mass and economic value. This allowed the environmental impacts associated with meat production to be reduced to account for the impacts of wool production.

Losses to the environment

A large number of nutrient loss pathways could be considered in the nutrient management assessment. Many of these loss pathways account for only small amounts of nutrients, or occur at infrequent intervals (i.e. bushfires). However, some of the loss pathways are essential for calculating other impacts (acidification, eutrophication) and these received more attention. After defining the pool of nutrient inputs to the system, the effect of major loss pathways could be estimated with a greater degree of confidence and the ability to verify assumptions. This is the context to the estimation of nutrient losses to the environment.

Soil/water loss

Losses of nutrients to the environment with eroded soil particles and overland flow can be a significant threat to the environment. Data for this section come from a detailed assessment of erosion losses and water quality impacts. The nutrient balance defines the pool of nutrients in the system that may be lost via different pathways, giving parameters to these estimates. For soil erosion, the nutrient loss pathway was quantified by the product of estimated soil erosion (see the soil erosion assessment) and estimated soil nutrient content. This provided a very broad estimate of nutrient loss, which could only be improved by detailed on-farm testing and research.

Water quality deals specifically with nutrient transfer to waterways. Within this assessment, data for the likely annual nutrient losses with runoff were collected (see Table 6 and Table 11). These data were cross checked with the nutrient balance (inputs less outputs). However, the nutrient loss off-paddock does not necessarily represent the amount of nutrient deposited in waterways, as discussed in the water quality section.


Page 14

Nutrient leaching

Leaching is the process of nutrient movement through the soil profile with water infiltration. Only nutrients that dissolve in the water rather than being bound in the soil structure can leach in significant quantities. While P and K can be subject to leaching, the greatest concern is usually associated with N leaching in the nitrate (NO32-) form. This can lead to groundwater contamination and acidification (see Table 4). K is also mobile in the soil solution and may leach at significant rates in some instances (i.e. Roberts 1970). However the available literature was limited and did not allow estimation of K losses in the systems considered.

Nitrate leaching was quantified by comparing the systems with data found in the literature. These data (see Table 4) are highly variable depending on experimental technique and location effects. Our model used these data as a guide to estimating leaching rates (which the model defines as a percentage of the total N input).

Table 4: Nitrate leaching rates under clover based pastures in southern Australia

Pasture type Research region

Soil Texture and annual rainfall

(mm/yr)

Nitrate leaching

(kg N/ha/yr)

Reference

Rye grass / white clover South eastern VIC Clay loam, 1114 mm 3.7 – 14.6

(Eckard et al. 2004)

Sub clover (annual) South western WA Sand, 460 mm 17 – 28

(Lundie et al. 2005)

Annual North eastern VIC Sandy clay loam over clay, 590 mm

82 (Ridley 1990)

Perennial North eastern VIC Sandy clay loam over clay, 590 mm

68 (Ridley 1990)

a The nitrate leaching rates in the experiment carried out by these authors were estimated from overall observations of pH decline over time rather than direct measurement.

Other factors influencing the selection of the leaching rate include the soil type, annual rainfall and rainfall pattern, pasture production N inputs to the system and other indicators of nitrate leaching (i.e. accelerated soil acidification). The leaching rate was also compared with the overall nutrient mass balance to check the effect of the assumptions used. Even a small difference in nitrate leaching may have a significant effect on other parameters, particularly soil acidification. However, collection of real data on nitrate leaching rates in Australia has been limited, and this limits the scope of nutrient budgeting in the context of this assessment. The data in Table 4 provide a guide to nitrate leaching for the systems analysed. Because of the higher rainfall and sandy loam soils at the Victorian site, leaching formed a significant component of the N losses at this site (30-34 kg/ha/yr), compared to those for NSW and WA (4-15 kg/ha/yr) where significantly lower rainfall was experienced in the years considered.


Page 15

Volatilisation

Volatilisation and denitrification refer to the loss of N from agricultural systems in a gaseous form (NH3, N2, NO). These losses can also have a deleterious effect on the environment through their role in the production of acid rain (NH3) and the greenhouse effect (nitrous oxides - NOx). Representative loss rates were sourced from the literature to provide some indication of the magnitude of these losses, but they are not considered in further detail. From the literature provided in Table 5, total gaseous N losses from the three systems was estimated to range from 20 – 27 kg/ha/yr depending on the level of N input to the system. Assessment of losses from the WA and NSW systems was difficult as similar systems are not covered as extensively in the literature. In the absence of other data it was assumed that N2 emissions were equal for all properties. It was assumed that because of the dryer climate in the NSW and WA system NOx losses will be lower (0.2 kg after Dalal et al. (2003)). It is recognised that the estimated N volatilisation rates are conservative with respect to environmental impacts and may be lower depending on specific management practices on farm. These values are included in the LCI for indicative purposes to highlight possible loss pathways only. Quantification of these nutrient losses can be very difficult even with experimental research, and the literature generally represents measurements made at a specific site over a relatively short time frame. These complexities reduce the degree of confidence in the model outputs for this category. Further improvement would require more research in the field of gaseous emissions from intensive and extensive grazing enterprises over time across different systems in Australia.

Table 5: Volatilisation and denitrification losses from agricultural systems in Australia

Pasture type Research region N2 losses

(kg N / ha)

NOx losses

(kg N / ha)

NH3 losses

(kg N / ha)

Reference

Rye grass / white clover

South eastern VIC 6 - 17


Rye / clover + 200 kg N as urea

South eastern VIC 13 - 57


Dairy pasture VIC - 6 – 11

(mean 8.5) - (Dalal et al. 2003)

Extensive pasture Australia wide 0.2

(Dalal et al. 2003)

Table 6 below shows the values used to calculate the nutrient budget for farms in the three supply chains. Data are presented as a range covering the three properties and two years of data collection.


Page 16

Table 6: Summary of nutrient output values and assumptions used for properties in the red meat supply chains

Nutrient outputs Description Value Reference Export of produce Sheep / Cattle

Wool

N = 2.4% of liveweight P = 0.7% of liveweight K = 0.2% of liveweight N = 16% of clean weight

QDPI&F (2005)

Improved pasture

N = 0.2 – 3.0 kg/ha/yr P = 0.1 – 2.5 kg/ha/yr

Extensive pasture N = 0.1 – 0.3 kg/ha/yr P = 0.02 – 0.1 kg/ha/yr

Nutrient loss with overland flow (N & P)

Cropping N = 0.2 – 6.0 - kg/ha/yr P = 0.1 – 2.0 kg/ha/yr

See Table 11 for a range of references

Soil loss rates by sheet and rill erosion and, gully erosion (used as input for nutrient losses)

Sheet and rill erosion = 0 – 5 t/ha/yr Gully erosion = 0 – 0.2 t/ha/yr

NLWRA (2001) Nutrient loss with soil erosion

Soil loss estimate multiplied by nutrient concentration within soil

N = 0.0 - 0.1 kg/ha/yr P = 0.0 - 0.1 kg/ha/yr

NLWRA (2001) and on farm soil analysis data

Leaching (N) N leaching dependant on soil type, rainfall and N mass balance

N= 4-34 kg/ha/yr (10-40% of N inputs from fertiliser and legumes)

See Table 4 for a range of references

Volatilisation (N) N2, NOx, NH3 N = 20-27 kg/ha/yr See Table 5 for a range of references

2.8 Soil Acidification Potential

Estimating the acidification potential in this LCA involved incorporating an acid / base balance into the model. This used published data for acidification rates resulting from product removal, N fertiliser acidification and grazed pasture acidification potential on improved or extensively managed pastures. The assumptions of the balance were matched as closely as possible to the known systems, considering the annual rainfall, soil type and the likely leaching potential, proportion of legume pasture, estimated annual dry matter production and grazing management system.

The model is based on five main drivers of acidification, balanced by the inputs of alkali into the system (livestock imports, some fertilisers, lime and other soil ameliorants). The drivers of acidification considered in the model are:

1. The use of N fertilisers. 2. Movement of agricultural products within the system through grazing animal

behaviour (transfer of manure to stock camps or laneways). 3. The use of legume-based pastures.


Page 17

4. The removal of agricultural products resulting in a net export of alkalinity. 5. Management that promotes a build up of soil organic matter.2

The impact of agricultural production on acidification was assessed by using the indicator 1 kg of calcium carbonate (CaCO3) equivalent. This indicator represents the amount of a substance, relative to calcium carbonate, required to neutralise the acidifying effect of a process. It is noted that this indicator of acidification potential does not represent the actual change in pH that may be observed from a management practice. This is because a range of factors affect pH change, particularly the buffering capacity of the soil. A summary of the assumptions and data used in the acidification estimates is provided in the following sections and in Table 10.

2.8.1 N fertiliser usage

The first form of acidification considered in the model comes from repeated applications of fertiliser N and particularly ammonium-based fertilisers (Bolan et al. 1991; Moody 2005). While it has been clearly demonstrated that there is no acidification effect from N transformations in a system with no N losses (Helyar 1976), this is rarely the case in practice (Bolan et al. 1991). N is lost from agricultural systems via several main pathways including product offtake, volatilisation and leaching of nitrate. Nitrate leaching was estimated as part of the nutrient management section and the figures produced were used to estimate acidification in this section of the model.

When nitrate (NO32-) leaches through the soil profile with a basic cation, H+ ions are deposited in the soil producing a net acidification effect. For each kmol of H+ remaining in the soil following nitrate leaching it is assumed that 50 kg of CaCO3 is required to neutralise the acidifying effect (Slattery 1991). Acidification rates (expressed as the amount of lime required to neutralise the effect) resultant from the application of some N fertilisers are presented Table 7 below.

Table 7: Lime required to neutralise the acidifying effects of some nitrogenous fertilisers at different rates of NO3 leaching

Fertiliser Lime requirement in kg CaCO3 / kg N applied Percentage of N applied leached (as nitrate) 0% 50% 100%

Ammonium sulfate 3.6 5.4 7.1 MAP 3.6 5.4 7.1 DAP 1.8 3.6 5.4 Nitram 0 1.8 3.6 Urea 0 1.8 3.6

Table adapted from (Moody 2005)

The net acidifying potential of ammonium based fertilisers results from surplus H+ ions being added to the soil following N transformations with or without nitrate leaching. These values are reported in the context of the Queensland broad-acre production

2 Soil organic matter can act as an acidifying substrate and as a buffer against pH change, resulting in different effects over time and between soil types.


Page 18

region. However the processes are expected to be similar in other regions. The major variable between regions affecting the acidification resultant from N application is expected to be the amount of nitrate leached from the system.

The model uses the leaching rate to calculate the acidification potential from fertiliser usage based on the data from Table 7 above. This was checked by calculating the kilomoles of N leached and the associated kilomoles of H+ ions left in the soil matrix following the theory outlined in Bolan et al. (1991). This theory was also used to estimate the acidifying effect of N leaching under legume pastures.

2.8.2 Grazing animal behaviour

The simple calculation of net acidification from product removal is complicated in grazing systems by the transfer of product within the paddock by disproportionate manure deposition across the grazing area. This effect can be attributed to the behaviour pattern of grazing animals, which graze over a large proportion of a paddock but select a small area on which to camp. These stock camps receive higher manure deposition resulting in nutrient transfer from the paddock to the stock camp. This can result in a 34% transfer of manure and urine to the stock camp area (Hilder, 1964, cited in Slattery 1991). While the addition of anions in manure (particularly calcium) could result in decreased acidification on the stock camp area, the higher rate of N addition and organic matter increases acidification on stock camp areas (Cayley et al. 2002).

The result of anion transfer to stock camping areas is acidification of the majority of the paddock where livestock graze. In addition to this, the research by (Cayley et al. 2002) indicates that net acidification also results at the stock camp site because of the higher deposition of N and organic matter. Considering this, the overall grazing effect results in a higher level of net acidification than the net removal of livestock may suggest. Table 8 below shows the acidification potential as a result of grazing behaviour in sheep.

The effect of grazing on acidification is influenced heavily by the management system on the property which affects the behaviour of livestock in establishing stock camp areas. Published data for the acidification potential attributable to sheep grazing are presented in Table 8. These data were used in the model where relevant to the supply chains.

Table 8: Potential acidification from sheep grazing behaviour

Species Research region Conditions Acidification

potential kg / ha / yr (CaCO3 equivalents)

Reference

Sheep South western VIC Perennial grass /

sub clover 9 – 25 (Cayley et al.

2002)

Sheep South Australia Extensive grazing 10 – 25 Ag Bureau of SA

Sheep North eastern VIC

Perennial grass / sub clover

23 (Slattery 1991)

* Numbers presented here are derived from the research of Slattery et al. (1991) (manure and urine acidification potential) and Hilder (1964) (Manure transport to stock camps).


Page 19

No data were available for the effect of cattle grazing on acidification. However anecdotal evidence suggests that cattle do utilise stock camps under extensive grazing conditions. Considering the lack of research for this effect with cattle production, a mean value from the sheep research referenced above was substituted where cattle are observed to display some camping behaviour on the supply chain properties. Grazing effects were estimated following subjective assessment of the supply chain properties. From this it was assumed that no grazing effect was evident on the Victorian property where cell grazing was the dominant pasture management practice. For the New South Wales and Western Australian properties, a value of 8-10 kg CaCO3 / ha / yr was selected as an approximation from the data presented in the literature.

2.8.3 Net removal of product

The transfer of agricultural products off-farm results in alkalinity exports that cause acidification. Table 9 summarises the values quantified by Moody (2005), NLWRA (2001) and Slattery et al. (1991) which underpin the acid / base balance used in the LCI model. In the situation where hay or grain was produced on-farm and then fed to livestock, the acidification effect from removing this product was still measured. This is because internal transfer of alkalinity may still produce acidification of grazing property land despite the produce not leaving the property.

2.8.4 Legume fixation and N leaching

N fixation from legume based pastures is an essential N source for grazing systems, as identified by the nutrient management section. However, excess N fixation from legume-based pastures can produce acidification from NO32- leaching below the root zone.

The nitrate leaching rate below annual and perennial legume based pastures has been studied by several research groups (Ridley et al. 2001; Eckard et al. 2004; Lundie et al. 2005) which reported a range of leaching rates for different pastures, soil types and rainfall patterns. The current research attempted to match the published leaching rates as closely as possible to the case study properties in the supply chains to determine the likely acidification potential. Leaching rates were determined from the nutrient balance, and the acidification potential from this rate of N leaching was determined using the acidification potential theory explained by Bolan et al. (1991).


Page 20

Table 9: Alkalinity in exported agricultural produce and lime requirement to neutralise acidifying effect of product removal

Product Unit* CaCO3

equivalent kg/ t of product

CaCO3 requirement (kg/ha) for some representative

yields

Reference

Wheat 1 t 9 18 (2 t/ha yield) (Slattery 1991)

Barley 1 t 8 16 (2 t/ha yield) (Slattery 1991)

Lupins 1 t 20 20 (1 t/ha yield) (Slattery 1991)

Grass hay 1 t 25 125 (5 t/ha yield) (NLWRA 2001)

Grass hay 1 t 30 150 (5 t/ha yield) (Moody 2005)

Clover hay 1 t 40 200 (5 t/ha yield) (NLWRA 2001)

Lucerne hay 1 t 60 300 (5 t/ha yield) (Slattery 1991)

Legume hay 1 t 50 250 (5 t/ha yield) (Moody 2005)

Sheep Meat 1 kg livewt 0.017 6 (10 x 35 kg lambs) (Slattery 1991)

Wool 1 kg 0.014 0.6 (5 kg / sheep x 8 sheep) (Slattery 1991)

*All values have been translated into standard units of 1 t or 1 kg for ease of comparison.

2.8.5 Increased soil organic matter

Increased soil organic matter levels are generally considered beneficial for soil health, structure and fertility. However, increasing the amount of organic matter cycling in a system by improving plant production or adding organic matter with manure can promote soil acidification (Sandars et al. 2003; Moody 2005). The acidifying process is driven by the dissociation of organic acids from the additional organic matter.

Research suggests that acidification may also result from additions of feedlot manure and effluent to pastures (Bouwman and Van Der Hoek 1997). However, manure can have a variable effect depending on its nutrient and organic matter content (Schoenau 2005). Some research reports decreasing pH from manure application (Chang et al. 1990), while other research (Whalen et al. 2000) reports an increase in pH following manure application on two acid soils under laboratory conditions.

Increased soil organic matter levels under improved pastures can contribute to soil acidification (Sandars et al. 2003), though there is a high degree of variability in the


Page 21

literature as to the extent of this effect (Crawford et al. 1994; Cayley et al. 2002; Sandars et al. 2003).

Organic matter was considered in the assessment of acidification as an influencing factor but not as a direct input. This is due to the lack of rigorous data to quantify the effect in isolation because of the confounding interactions in pasture systems. It was assumed that acidification as a result of organic matter additions was far lower than other acidifying processes and the omission is not expected to cause significant error.

2.8.6 Summary of Soil Acidification Data

Table 10 below summarises the values estimated for the soil acidification balance. The data have been presented as a range covering the three properties and two years of data collection.

Table 10: Summary of acidification potential data used for properties in the supply chains

Acidifying Process Description CaCO3 required for Neutralisation

References

N leaching from fertiliser usage

Acidification depends on percentage of N leached (10-40% from nutrient balance)

0 – 6 kg/ha/yr averaged across the whole property

Estimates based on data from Moody (2005) and Slattery (1991)

Animal Grazing Behaviour

Acidification caused by transfer of alkalinity within paddocks to livestock camping areas

0 kg/ha/yr for the VIC supply chain. 10 kg/ha/yr for NSW and WA. Averaged across the whole grazing area

See Table 8

Net product removal Acidification caused by removal of alkalinity with plant and animal products


See Table 9

N leaching from legume pasture

Acidification depends on percentage of N leached (10-40% from nutrient balance).

VIC = 106-121 kg/ha/yr NSW = 13-51 kg/ha/yr WA = 24-45 kg/ha/yr Data averaged across the whole property

Lundie et al. (2005) Eckard et al. (2004) Ridley et al. (2001) Bolan et al. (1991)

Alkalinity Additions

Description

CaCO3 added (kg/ha/yr)

References

Lime and soil additives

Lime added to ameliorate low soil pH

0 kg / ha / yr – NSW and WA supply chains. 1177 kg/ha/yr with lime application on VIC supply chain property in 2002 . Values averaged across the whole property.

-

Net product inputs to property (Hay, grain, livestock)

Main imports are livestock and hay


See Table 9 for composition data.


Page 22

2.9 Soil Erosion

The assessment of erosion for the red meat supply chain properties was based on broad scale erosion estimates sourced from the NLWRA (2001) erosion research. These estimates all relate to water borne erosion and no assessment of wind erosion was considered, although this will be important for future assessments of northern Australian beef production. The erosion mapping was ground-truthed by considering the topography and management system used on each supply chain property. As erosion is a natural process in the Australian landscape, the rate of erosion for each property was compared to natural or pre-European erosion rates calculated by the NLWRA (2001). This reduced the risk of attributing natural erosion processes to red meat production.

The land on which the NSW property is situated has an estimated erosion gully density of 0.1 to 0.5 km/km2, described as “low density”, while the estimated annual hillslope erosion rate ranges from 0.5 to over 10 t/ha/yr (“low” to “very high”). This reflects the soil types and topography of the area. At the Victorian property, the estimated erosion gully density is 0 to 0.1 km/km2, described as “very low” density, and the estimated annual hillslope erosion rate ranges from 0 to 0.5 t/ha/yr (also “very low”) which is approximately equal to pre-European erosion rates for this area. Between these two estimates, the WA property has an estimated erosion gully density of 0 to 1 km/km2, described as “very low” to “medium” density, while the estimated annual hillslope erosion rate ranges from 0 to over 2.5 t/ha/yr (“very low” to “low”).

The model estimates erosion using these NLWRA soil loss mapping data and a subjective assessment of the factors likely to accelerate soil erosion on each of the supply chain properties. This is intended to provide an indication of the erosion risk on land used for red meat production in different regions across the country. It was beyond the scope of this project to assign soil loss to the functional unit (1 kg of HSCW) as this would imply a definite link between meat production and erosion. Assessment of soil erosion risk in the red meat industry would ideally involve establishing new indicators related to soil disturbance from livestock trampling and reduction in vegetative cover from grazing. At this point there is insufficient quantifiable research into the relationship between these processes and erosion potential to establish these indicators. While there are some linkages related to the modification of ground cover through grazing management and land clearing and trampling, these are only partly responsible for accelerated erosion. Further background on erosion as it relates to the red meat industry may be found in Wiedemann et al. (2006).

The model provides an estimate of soil loss from the supply chains on a per hectare basis using a second functional unit ‘1 ha of land used for production’ to identify the magnitude of the effect. This avoids attributing an NRM impact that is related to many factors extraneous to the production of red meat to this one system output.

2.10 Water Quality

Water quality is addressed in the red meat LCA through eutrophication potential by estimating exports of N and P to waterways from red meat supply chain properties. Nutrient loss results from the dissolution of readily available nutrients in overland flow, and the loss of soil and organic matter particles eroded with overland flow and carried to waterways. While there is literature available to quantify nutrient losses from small plot


Page 23

scale experiments, it is not always accurate to assume similar nutrient export rates at the whole property scale (Barlow et al. 2005). This current research is based on a desktop study without on-farm research. It attempts to provide a broad-scale estimate of nutrient loss from the supply chain properties. Nutrient loss at the property scale has been researched by as few as five studies in Australia (Barlow et al. 2005) and none of these are representative of the systems being considered in the red meat LCA. Hence the model uses conservative estimates of nutrient export while matching these as closely as possible to the systems being studied.

2.10.1 Nutrient loss in overland flow

Nutrient loss is a measure of concentration of nutrient and volume of runoff. Hence the model attempted to account for differences in rainfall between regions and across years. Several researchers have reported data for estimated nutrient loss with overland flow (Greenhill et al. 1983; Nash and Murdoch 1997; Nash and Halliwell 1999; Ridley et al. 2003; Barlow et al. 2005; Nash et al. 2005). These data vary widely, as seen by Table 11 below.

Table 11: Nutrient losses from pasture systems is Australia

Pasture type Research region

N conc. in runoff

(mg/L - NO3-N)

N losses

(kg N/ha)

P conc. in runoff

(mg/L)

P losses

(kg P/ha)

Reference

Control pasture (no fert inputs)

Westernport VIC


Page 24

2005). This may be because of the catchment of water in on-farm dams and the filtration of nutrients out of solution by riparian zones. Variations within paddocks or between different runoff events were not considered in developing estimates of nutrient loss from properties in the supply chains as data were not available to quantify this variation. The reference literature was used to provide an estimate of nutrient runoff per hectare per year. A conservative approach to nutrient loss has been taken because of the diluting effect expected when assessing a whole farm compared to literature presented on a paddock scale (after (Barlow et al. 2005).

Nutrient loss per hectare is influenced by the level of soil fertility and the amount of nutrients at the surface of the soil. Intensively grazed areas have a higher nutrient turnover and higher deposition of manure on the soil surface and hence higher nutrient levels in runoff. At the Victorian site (intensively grazed), average losses of 3 kg N / ha / yr (Ridley et al. 2003) and 2.5 kg P / ha / yr (Barlow et al. 2005) were estimated for intensively grazed areas. Losses at the NSW site were assumed to be 3 kg N / ha / yr and 0.45 kg P / ha / yr because of the lower intensity of the grazing system (Ridley et al. 2003). At the WA site, N losses of 0.2 kg N / ha / yr were estimated (Ridley et al. 2003), while P losses of 0.1 kg P / ha / yr (Costin 1980) were estimated in response to the low soil fertility at this site and lower grazing intensity at this site. Nutrient losses from cropped areas were also relevant at the NSW and WA sites. At the NSW site losses off the alluvial cropping areas were estimated to be 6 kg N / ha / yr and 2 kg P / ha / yr, while estimates for cropping areas in WA were 0.2 kg N / ha / yr and 0.1 kg P / ha / yr reflecting the lower soil fertility at this site.

2.10.2 Nutrient loss with soil and organic matter erosion

Erosion of soil and organic matter particles results in nutrient loss. The model estimated the magnitude of nutrient losses through erosion as the product of soil loss estimates (soil erosion sheet) and the estimated soil nutrient content. As limited data on soil nutrient levels were available for the supply chain properties considered, the values were estimated based on property management and production. These assumptions may introduce error into the evaluation of eutrophication. Also, the estimate of soil erosion rates was made from broad-scale data

comp.094 final report...southern red meat production – a life cycle assessment page iv the...

Documents